Melt pool height monitoring based on dynamic laser spot in Oscillating Laser Arc Hybrid Additive Manufacturing

Oscillating Laser Arc Hybrid Additive Manufacturing (OLAH-AM) combines the deposition efficiency of arc with the energy-manipulation capability of an oscillating laser, making it suitable for fabricating large-scale metal components. A key challenge in achieving automation and repeatability in OLAH-AM is the lack of robust technologies capable of reliably monitoring melt pool height. The fundamental advancement of this study is a novel melt pool height measurement method that requires only a single co-planar camera. The innovation lies in using the laser spot that is inherently generated by the oscillating laser-arc hybrid heat source on the melt pool as the sensing target, enabling stable and rapid extraction of height information. The core of this method lies in accurately localizing the laser spot under complex melt pool interference. To this end, an improved object detection model was developed by incorporating a P2 feature layer and a weighted bidirectional feature pyramid network (BiFPN), which achieved a recall rate of 85.5 % and mAP of 93.3 %. The subsequent height fitting was accomplished by employing the Random Sample Consensus (RANSAC) algorithm for parabolic fitting and applying perspective transformation to the curve's peak point. The proposed method achieved a height measurement error of 3.75 %, with each measurement requiring only 41.37 ms. Validation experiments confirmed the effectiveness of this approach, which lays a crucial foundation for future real-time stability control in robotic OLAH-AM processes.